Sponsored content

By Anne Hunter, VP of Product Marketing, DISQO

Surveys are a strong and proven tool for understanding consumer motivations, interests and opinions. However, the limits of human recall and self-censorship make them imperfect instruments to gather data on people’s behaviors. When respondents are asked to report on things that occurred in the past, they often fail to be 100% accurate. In addition, when predicting what they will do going forward, they may misjudge their future actions. Can today’s busy and overstimulated consumer really remember everything that happened in the past and know with certainty how their days will be spent in the future?

The variance between what people think or say and what actions they actually take is known as the say-do gap.

A clear example of this gap emerged during a study we conducted on the opinions and behaviors of members of a politically oriented social network. Leveraging zero-party data from our large 100% opt-in U.S. behavioral audience panel, we randomly selected a sample of survey respondents who had visited the social platform and assigned them to test and control groups. The test group was first asked who they voted for in the last election and then were asked about the websites they had recently visited. The control group was first asked what type of car they owned and then asked the same series of questions about media consumption.

As expected, neither group was able to accurately describe their past media behaviors but the control group exhibited a higher accuracy in describing their media consumption compared to the group asked about their political preferences. While it’s long been understood that survey question sequencing and framing can have an impact on attitudinal responses, it is notable that we detected a difference in the way that initial question frame impacted the accuracy of self-reported behavior. There was indeed a gap between how consumers report their online behavior in a survey and how they actually behave online, which can be further exacerbated by the survey instrument itself.

Surveys of consumers can provide an excellent understanding of the “why.” However, bringing survey recall and passively collected behavioral data together from a single source is what truly closes the gap because different things can actually happen than what people originally intended or wanted.

In market research, understanding the say-do gap becomes critical to the success of corporate initiatives in several ways:

Brand equity tracking: Among the myriad benefits that brand tracking provides, adding behavioral data can catch early signals of change and create better predictions of sales performance than survey data alone.

Marketing strategy: Leveraging behavioral data in tandem with survey data uncovers a complete, holistic view of the customer journey. Connecting people’s opinions to their actual behaviors is key in understanding how to effectively market to consumers at every touchpoint. 

Advertising effectiveness: Not all media can be easily measured through survey tags. By adding behavioral panel views of actual ad exposure, such as on social media sites, a completely distinct and neutral analysis of advertising performance can be executed for media optimization, including the impact on sales behaviors.

Product development and innovation: Rich consumer insights are vital at every product stage from ideation and testing to commercialization and beyond. Identifying and closing the say-do gap drives new product innovation and ensures product success.

Customer experience: People at the extremes of experience are the most likely to respond to CX feedback mechanisms. Analyze actual experience for the general population instead of only hearing from the most vocal customers by combining survey feedback with passive behavioral data to achieve a true understanding of the drivers of experience.

As the need for accurate insights continues to grow, researchers must determine when to apply a standalone survey-driven methodology, when to leverage observed behavioral data or when a combination of the two would be best. Have you checked your insights for the say-do gap?

www.disqo.com